Superhydrophobic diving flies (
Ephydra hians
) and the
hypersaline waters of Mono Lake
Floris van Breugel
a,1
and Michael H. Dickinson
a
a
Department of Biology, California Institute of Technology, Pasadena, CA 91125
Edited by Jerrold Meinwald, Cornell University, Ithaca, NY, and approved October 18, 2017 (received for review August 22, 2017)
The remarkable alkali fly,
Ephydra hians
, deliberately crawls into the
alkaline waters of Mono Lake to feed and lay eggs. These diving flies
are protected by an air bubble that forms around their superhydro-
phobic cuticle upon entering the lake. To study the physical mecha-
nisms underlying this process we measured the work required for
flies to enter and leave various aqueous solutions. Our measure-
ments show that it is more difficult for the flies to escape from Mono
Lake water than from fresh water, due to the high concentration of
Na
2
CO
3
which causes water to penetrate and thus wet their setose
cuticle. Other less kosmotropic salts do not have this effect, suggest-
ing that the phenomenon is governed by Hofmeister effects as well
as specific interactions between ion p
airs. These effects likely create
a small negative charge at the air
–
water interface, generating an
electric double layer that facili
tates wetting. Compared with six
other species of flies, alkali flies
are better able to resist wetting in
a0.5MNa
2
CO
3
solution. This trait arises from a combination of
factors including a denser layer of setae on their cuticle and the
prevalence of smaller cuticular hy
drocarbons compared with other
species. Although superbly adapte
d to resisting wetting, alkali flies
are vulnerable to getting stuck in natural and artificial oils, including
dimethicone, a common ingredient in sunscreen and other cosmetics.
Mono Lake
’
s alkali flies are a compelling example of how the evo-
lution of picoscale physical and chemical changes can allow an ani-
mal to occupy an entirely new ecological niche.
superhydrophobicity
|
insects
|
Hofmeister series
|
bubbles
|
biomechanics
I
n late summer the shores of Mono Lake, California, are bus-
tling with small flies,
Ephydra hians
, which crawl underwater to
feed and lay eggs (Fig. 1
A
). Their unusual behavior was eloquently
described by Mark Twain during his travels to Mono Lake (1):
You can hold them under water as long as you please
—
they do not
mind it
—
they are only proud of it. When you let them go, they pop up
to the surface as dry as a patent office report, and walk off as un-
concernedly as if they had been educated especially with a view to
affording instructive entertainment to man in that particular way.
Although Twain
’
s observations are over 150 years old, we still
do not understand the chemistry and physics underlying the ability
of these flies to resist wetting as they descend below the water
surface. Alkali flies are found on nearly every continent and fulfill
an important ecological role by transforming the physically harsh
environments of alkaline lake shorelines, including the Great Salt
Lake in Utah and Albert Lake in Oregon, into important wildlife
habitats (2). Aside from the flies, only algae, bacteria, and brine
shrimp tolerate Mono Lake
’
s water, which is three times saltier
than the Pacific Ocean and strongly alkaline (pH
=
10) due to the
presence of sodium bicarbonate and carbonate. For the past
60,000 y, Mono Lake has had no outlet (3), driving a steady in-
crease in the concentration of mineral salts through a yearly
evaporation of 45 inches (4). Calcium from natural springs un-
derneath the lake
’
s surface reacts with the carbonate-rich water,
precipitating calcium carbonate in the form of underwater towers
called tufa. Alkali flies crawl underwater by climbing down the
surface of the tufa, which have become exposed due to falling lake
levels (Fig. 1
B
and
Movies S1
and
S2
).
For the alkali fly, staying dry is paramount to their survival; if
they do get wet in Mono Lake, a thin film of minerals dries on their
cuticle, which makes them more likely to be wetted in subsequent
encounters with the water. Like most insects, the flies are covered
in a waxy cuticle festooned with tiny hairs (setae). As in water
striders (5), these hydrophobic hairs trap a layer of air, so that as a
fly crawls into the water an air bubble forms around its body and
wings. The bubble protects the flies from the salts and alkaline
compounds present in the lake and also serves as an external lung
(6), allowing flies to spend up to 15 min underwater crawling to
depths of 4
–
8 m (7). Once finished feeding or laying eggs the flies
either crawl to the surface or let go of the substratum and float up.
As noted by Twain (1), the bubble pops when it hits the air
–
water
interface, depositing its inhabitant safe and dry on the water
’
s
surface (Fig. 1
C
and
Movies S3
and
S4
). In this paper we describe
the physical and chemical properties that make the alkali flies
uniquely able to form these protective bubbles in Mono Lake
’
s
dense and alkaline waters.
As a preamble to our measurements, we briefly review the
physics of solid
–
liquid interactions. On smooth surfaces, the
shape of an adhering liquid droplet may be described by the con-
tact angle, with larger contact ang
les corresponding to less-wettable
surfaces (Fig. 1
D
). The contact angle for a smooth piece of waxy
insect cuticle is typically 100
–
120°, similar to paraffin wax, and close
to the theoretical maximum (8, 9). On rough surfaces, like that of
an alkali fly, a liquid drop can exist in two different states. In the
Cassie
–
Baxter state, air pockets fill the space between roughness
elements (10), resulting in
“
superhydrophobicity
”
[a.k.a. the
“
lotus-
effect
”
(11)]. In the Wenzel state, the liquid replaces the air
pockets (12), resulting in a fully wetted surface. Whether a liquid
–
surface interface exists in the Cassie
–
Baxter or Wenzel state is a
complex function of the surface
’
s physical and chemical structure,
Significance
Superhydrophobic surfaces have been of key academic and
commercial interest since the discovery of the so-called lotus
effect in 1977. The effect of different ions on complex super-
hydrophobic biological systems, however, has received little
attention. By bringing together ecology, biomechanics, physics,
and chemistry our study provides insight into the ion-specific
effects of wetting in the presence of sodium carbonate and its
large-scale consequences. By comparing the surface structure
and chemistry of the alkali fly
—
an important food source for
migrating birds
—
to other species we show that their uniquely
hydrophobic properties arise from very small physical and
chemical changes, thereby connecting picoscale physics with
globally important ecological impacts.
Author contributions: F.v.B. and M.H.D. designed research; F.v.B. and M.H.D. performed
research; F.v.B. and M.H.D. contributed new reagents/analytic tools; F.v.B. analyzed data;
and F.v.B. and M.H.D. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
Published under the
PNAS license
.
Data deposition: All data in the manuscript have been uploaded to Github (
https://github.
com/florisvb/alkali_flies_of_mono_lake
) and Open Science Framework (
https://osf.io/
43yhs/
).
1
To whom correspondence should be addressed. Email: florisvb@gmail.com.
This article contains supporting information online at
www.pnas.org/lookup/suppl/doi:10.
1073/pnas.1714874114/-/DCSupplemental
.
www.pnas.org/cgi/doi/10.1073/pnas.1714874114
PNAS
|
December 19, 2017
|
vol. 114
|
no. 51
|
13483
–
13488
ECOLOGY
CHEMISTRY
the chemistry of the solution, and the interactions between the
surface and the liquid. In the case of an alkali fly crawling into
water, the combination of hydrophobic wax and setose surface fa-
vors the Cassie
–
Baxter state, rendering the flies superhydrophobic
(6, 9, 13), with contact angles appr
oaching 180°. Other insects that
make air
–
water transitions, including
spiders and beetles, sport
small patches with superhydrophob
ic properties used for plastron
respiration (6).
Results
To investigate which chemical and physical properties of the flies
and Mono Lake water (MLW) influence the formation of the air
bubble that protects them from the mineral-rich water we built
an optical force sensor (Fig. 2
A
), which we used in a manner
similar to the Wilhelmy balance method (14) to measure the
forces required for flies to enter and exit different solutions. We
glued the flies to a tungsten beam (0.26-mm diameter) and
slowly submerged them using a linear motor (speed 0.3 mm
·
s
−
1
).
The average peak force required for flies to enter MLW was
∼
1 mN, roughly 18 times the body weight of the 5.5-mg flies
(Fig. 2
B
and
Fig. S1
A
). The force required to enter the water
varied with body orientation, with a minimum at a vertical,
headfirst orientation (
Fig. S1
B
). This corresponds with our ob-
servations of flies at Mono Lake, which tend to enter the water
by crawling down 45°
–
90° surfaces.
In pure water, the work required to submerge the fly is largely
recovered when it is pulled out of the water
—
the surface tension of
the bubble stores the potential energy much like a spring. Thus, we
use the term
“
recovered work
”
(Fig. 2
C
) as a measure of how easy
it is for the flies to escape the water. A positive value indicates a net
upward force that pushes the fly
out of the water, whereas a neg-
ative value indicates that the fly is partially wetted and trapped by
surface tension at the air
–
water interface. With increasing con-
centration of MLW from 0 to 200% we found that the recovered
work decreases, despite the incr
ease in solution density (Fig. 2
D
).
MLW contains a number of salts including NaCl, Na
2
SO
4
,and
K
2
SO
4
, as well as the alkali components sodium bicarbonate and
boric acid (15). To determine which of these components most
influences the recovered work we made two solutions, each con-
taining double the natural concentration of either the salts or al-
kali compounds. Sodium bicarbonate is an alkali buffer in which
the ratios of CO
3
2
−
,HCO
3
1
−
,andH
2
CO
3
are coupled to pH
according to the Henderson
–
Hasselbalch equation. To achieve a
pH equal to that of MLW (pH
=
10), we used a molar ratio of
NaHCO
3
to Na
2
CO
3
of 0.8. Compared with MLW, recovered
work was higher for the salt solution, whereas it was significantly
lower for the alkali solution (Fig. 2
E
), implying that the alkali
compounds make it more difficult for the flies to escape the water.
Fig. 1.
Mono Lake
’
s alkali flies must exert up forces to 18 times their body
weight to crawl underwater to feed and lay eggs. (
A
) Close up of an alkali fly
under water. (
B
) Image sequence of a fly crawling into the water (
Movies S1
and
S2
). (
C
) Image sequence of a fly floating upward to the surface inside its
air bubble (
Movies S3
and
S4
). (
D
) Illustrations of a water droplet on smooth
and rough surfaces.
Fig. 2.
High concentrations of sodium carbonate make it more difficult for
flies to escape from MLW. (
A
) Diagram of optical force sensor. Forces on the
fly deflect the beam, shifting the shadow cast by an LED, which is detected
by a photo detector. (
B
) Force vs. distance traveled (normalized to fly height)
for 20 CO
2
anesthetized flies dipped into MLW (bold: mean). Positive values
correspond to upward forces (
Movie S5
). (
C
) One example trace from
A
.
Shading indicates the amount of work done on the fly as it exits the water,
which we term recovered work. (
D
) Recovered work for different concen-
trations of MLW, ranging from pure deionized water to double-strength
MLW (produced via evaporation): 0% (
n
=
30), 50% (
n
=
20), 100% (
n
=
50),
200% (
n
=
20). Black line: expected recovered work based on the mea-
surements in pure water and the increase in solution density. Red line: data
regression (
P
=
0.003;
r
2
=
0.08). (
E
) Recovered work for salt or alkali solu-
tions. Salt solution (double concentration of the salts in Mono Lake): 1.3 M
NaCl, 0.2 M Na
2
SO
4
, 0.04 M K
2
SO
4
, and 1.8 mM K
3
PO
4
. Alkali solution
(double concentration of the alkali compounds in Mono Lake): 0.35 M
NaHCO
3
,0.44MNa
2
CO
3
, and 0.09 M boric acid. (
F
) Recovered work for a
0.5 M NaHCO
3
buffer solution at three different pH values. (
G
) Same as
F
,
showing only the subset of 10 flies that were dipped in the order of in-
creasing pH. (
H
) Recovered work for standard MLW and MLW neutralized to
pH 7 with HCl. (
I
) Recovered work for pure water and 5 mM NaOH, at the
same pH as the carbonate buffer in
E
and
F
.(
J
) Recovered work for HCl-
neutralized 0.5 M carbonate buffer, 0.16 M carbonate buffer, and 0.5 car-
bonate buffer. In this experiment all flies were dipped in the order from left
to right. (
D
–
J
) Shading indicates bootstrapped 95% confidence intervals.
Nonoverlapping shading generally corresponds to statistical significance of
P
<
0.02. Resampling test statistics are given for cases in which differences are
not obvious.
D
–
G
and
H
–
J
come from two separate collections, which may
explain the slightly higher water repellency in
H
–
J
than would be expected
based on
D
. The order of solutions into which the flies were dipped was al-
ternated for each set of experiments unless otherwise noted.
E
–
J
,
n
=
20. The
Bond number for flies in the 0.5 M Na
2
CO
3
solution is 0.46, indicating that
surface tension forces are dominant (
Supporting Information
).
13484
|
www.pnas.org/cgi/doi/10.1073/pnas.1714874114
van Breugel and Dickinson
Next, we tested three sodium bicarbonate buffer solutions ranging
in pH from 8.5 to 11.6 and found that high pH significantly de-
creased recovered work (Fig. 2
F
and
G
).
Our results with the bicarbonate buffer solution suggest that the
naturally high pH of the lake makes it more difficult for the flies to
escape the surface. To directly test this hypothesis we neutralized
MLW with HCl, bringing the pH down to 7, which should slightly
shift the carbonate balance toward HCO
3
1
−
. Compared with
natural MLW this neutralized solution did not significantly in-
crease the recovered work (Fig. 2
H
). Testing a different alkali
solution (5 mM NaOH) at the same pH as the highest-pH sodium
carbonate buffer (11.6) further confirmed that pH alone does not
determine the amount of recovered work (Fig. 2
I
). We next tested
the possibility that the concentration of Na
2
CO
3
played a critical
role by measuring the recovered work in three solutions: 0.5 M
Na
2
CO
3
(pH 11.6), 0.15 M Na
2
CO
3
(pH 11.6), and 0.5 M Na
2
CO
3
neutralized with HCl to pH 7. We found the recovered work was
lowest for the 0.5 M Na
2
CO
3
solution (Fig. 2
J
), which together
with the buffer experiments from Fig. 2
F
and
G
indicates that a
high concentration of Na
2
CO
3
makes it more difficult for the flies
to escape the water surface.
To test whether the alkali flies possess a unique adaptation to live
in Na
2
CO
3
-rich waters we compared their ability to recover work in
distilled water, MLW, and a 0.5 M Na
2
CO
3
solution to that of six
dipteran species, including two other members of the Ephydridae
(shore flies), two coastal kelp flies (adapted to living under constant
salty ocean spray), and two cosmopolitan drosophilids (Fig. 3
A
). All
species were similar to alkali flies in that the work recovered from
distilled water scaled with body length (Fig. 3
B
), which is expected
because surface tension forces on a
floating object are a function of
contact perimeter (16). However, work recovered from MLW and
the Na
2
CO
3
solution was significantly lower in the other species. Of
all species tested only the alk
ali flies were pushed out of the
Na
2
CO
3
solution, suggesting that they
have unique adaptations that
render them superhydrophobic in the presence of Na
2
CO
3
.
To investigate the physical differences between flies, we im-
aged samples of each species with SEM. The alkali flies possess a
denser mat of setae on their bodies and legs (Fig. 3
C
and
D
and
Fig. S2
) and lack obvious pulvilli between their tarsal claws
(Fig. 3
E
). In observing the other fly species in our plunging ex-
periments it is clear that the presence or absence of pulvilli does
not play a critical role. In trials in which recovered work is neg-
ative the entire body was wetted, not just the tarsi (Fig. 3
F
and
Movies S8
and
S9
). To quantify the hairiness of each species, we
used image processing to calculate the number of hair crossings
per micrometer for SEM image transects perpendicular to the
mean hair orientation (
SI Methods
). Based on these metrics, the
alkali flies are generally hairier than the other species: wings
(
+
34%), thorax (
+
44%), abdomen (
+
47%), tarsi (
+
17%), and
overall average (
+
36%). However, they are only 15% hairier than
Fucellia rufitibia
(a kelp fly). To summarize, body length explains
57% of the variance in recovered work in pure water across all
seven species (65% if alkali flies are excluded) but only 0.009% for
a0.5MNa
2
CO
3
solution (but 57% if alkali flies are excluded).
Fig. 3.
Mono Lake
’
s alkali flies are uniquely adapted to withstand wetting in alkali water. (
A
) For each of seven species we measured the recovered work for
pure water, MLW, and a 0.5 M Na
2
CO
3
solution. Only the alkali flies were actively propelled out of the carbonate solution; all other species were stuck at the
surface. See
Movies S6
and
S7
. (Scale bars, 1 mm.)
n
=
10 for Fr, Cv, Dm, Dv;
n
=
20 for Eh;
n
=
5 for Hp, Esp. (
B
) Correlation between body length and recovered
work from
A
. Alkali flies are indicated by a star and were omitted from the regressions. Shading indicates 95% confidence interval for the slope of the re-
gression. (
C
and
D
) SEM images of the thorax and tarsi for each fly species. (Scale bars, 10
μ
m.) (
E
) The alkali fly is unique among the species examined in its lack
of pulvilli. SEMs of three other representative species are shown (see also
Fig. S7
). (Scale bars, 10
μ
m.) (
F
) Cv before and after being dipped into 0.5 M Na
2
CO
3
.
van Breugel and Dickinson
PNAS
|
December 19, 2017
|
vol. 114
|
no. 51
|
13485
ECOLOGY
CHEMISTRY
After removing the body-size trend from the data for the Na
2
CO
3
solution, a positive correlation between hairiness and recovered
work explains 58% of the remaining variance.
To determine whether the flies
’
cuticular hydrocarbons might
act in combination with the setae to prevent wetting we briefly
rinsed flies in hexane and measured the recovered work in MLW
and distilled water. We found that hexane removed compounds
that are important for the fly to stay dry in MLW but not in pure
water (Fig. 4
A
). Next, we analyzed the cuticular hydrocarbons of
all seven species with GCMS (see
SI Methods
for details). The
cuticular hydrocarbon profile of alkali flies is dominated by
straight-chain alkanes (pentacosane [C25] and heptacosane [C27])
(Fig. 4
B
). The two other members of Ephydridae were similar,
whereas the two drosophilids had a higher abundance of larger
alkenes, dienes, and methylated even-numbered hydrocarbons.
The kelp flies exhibited very different profiles dominated by tetra-
methylated C30 and C21 (Fig. 4
C
).
To verify the reproducibility of our results, we developed a
simplified assay to test the effects of different solutions on flies
’
ability to escape from a liquid
–
air interface. We used the easily
reared species
Drosophila virilis
for these experiments because
they require large numbers and we did not wish to kill so many
wild-caught
Ephydra
. In these trials we briefly anesthetized
20 flies with CO
2
and sprinkled them onto a 44-cm
2
surface of
nine solutions, each at five concentrations. Sodium carbonate
was the most detrimental to the flies
’
ability to escape compared
with other salts (in particular at intermediate concentrations),
even compared with solutions of pH
>
13, those containing di-
valent anions, or K
2
CO
3
(Fig. 5). The small effect of K
2
CO
3
compared with Na
2
CO
3
suggests that the enhanced wetting
caused by Na
2
CO
3
is not solely a property of the carbonate ion
but also of its interaction with the sodium ions.
Our working hypothesis is that the presence of Na
2
CO
3
biases
the liquid
–
cuticle interaction to favor Cassie
–
Baxter-to-Wenzel-
state transitions. To try to observe this phenomenon more di-
rectly we developed another preparation that makes use of the
long fine hairs found on the trailing edge of many insect wings.
Wetting, or its absence, is easy to visualize when this 2D array of
hairs is placed in contact with a water drop. For these experi-
ments, we chose the common house fly,
Musca domestica
, due to
its large size and availability. We directly filmed the interaction
between the wings and drops of either pure water or 0.5 M
Na
2
CO
3
solution. In only one of nine wings did a tiny droplet of
pure water stick to the wing (
Movie S8
). In the case of 0.5 M
Na
2
CO
3
, however, four of the nine wings showed large drops
adhered to the wing, and one with a tiny droplet (
Movie S9
). The
influence of Na
2
CO
3
might act directly on the cuticle surface, or
it might involve a more complex mechanism involving geometry
of the fine hairs and the spaces between them. To test between
these possibilities we needed a sufficiently large piece of flat
chitin on which we could accurately measure contact angles.
Because insects are too small and setose, we made clean, flat
preparations from shrimp exoskeletons for these tests. We mea-
sured no difference in the contact angle for water and 0.5 M
Na
2
CO
3
[water: 81
±
14° (mean
±
SD), carbonate: 76
±
15°;
n
=
18 each;
t
test:
P
=
0.31,
t
stat
=
−
1; 2-
μ
L static sessile drop
technique; see
SI Methods
]. Although shrimp cuticle lacks the
hydrocarbons found on insects, chitin and hydrocarbons have
roughly similar surface free energies (17, 18), and thus Young
’
s
equation predicts that they will have similar contact angles as
well (14, 19). These results suggest that Na
2
CO
3
acts to favor the
Wenzel state by a mechanism involving the fine air pockets
between hairs.
In the course of our field work we frequently observed large
numbers of flies that were wetted and drowned on the surface of
Mono Lake. We hypothesized that such events were due to oils
from decaying organic matter that made it more difficult for flies
to escape the water. When dropped onto MLW coated in a thin
film of fish oil (20
μ
Lover44cm
2
) alkali flies immediately became
trapped on the surface like birds in an oil spill (
Fig. S3
A
and
B
and
Movies S10
and
S11
). In addition, while collecting water for
our experiments we occasionally noticed a thin film of sunscreen
coming off of our skin and considered whether this might also
deleteriously influence hydrophobicity. We measured the forces
on alkali flies dipped into untreated MLW and MLW used to
rinse our hands 5 and 15 min after applying sunscreen. The sun-
screen indeed had a catastrophic effect on the flies
’
ability to stay
dry (Fig. 6
A
–
C
). To examine this effect in more quantitative
detail we applied measured amounts of sunscreen to wooden
applicator sticks and stirred them in pure water for 1 min. After
briefly anesthetizing them with CO
2
we dunked the flies un-
derwater and scored each fly after 15 min for either having flown
away or gotten stuck. Amounts of 8
–
40 mg (applied to the wooden
sticks) of the Neutrogena Ultrasheer SPF 50 Sport sunscreen
raised the fraction of trapped flies from 50 to 100% (Fig. 6
D
). We
then tested 20-mg applications of six different brands and found
Fig. 4.
Increased hairiness and a coating of C25 cuticular hydrocarbons help
the alkali flies resist wetting in MLW. (
A
) Recovered work for pure water and
MLW before and after alkali flies were rinsed in hexane for three 1-s ses-
sions. Flies were dipped in one of two orders (
i
): deionized water, MLW,
hexane treatmentand a final dip in MLW (
n
=
10) or (
ii
) MLW, deionized
water, hexane treatment, and a final dip in deionized water (
n
=
10). (
B
) GC-
MS analysis of hexane extracted cuticular hydrocarbons of the alkali fly.
(
C
) Relative abundance of hydrocarbons found by GC-MS in hexane extracts
of each species; average of the retention times (in minutes) for all of the GC-
MS peaks (weighted by relative abundance); mean number of hairs per
micrometer (averaged across thorax, abdomen, wings, and tarsi); approxi-
mate body length of the species, in millimeters; and subjective relative size
of the pulvilli. Codd: odd-length straight-chain hydrocarbons (e.g., C25
and C27). MeCodd: methylated odd-length carbon chains (e.g., 3Me-C25).
Alk/Die Codd: odd-length carbon chain alkenes and dienes. Ceven: even-
length straight-chain hydrocarbons (e.g., C26 and C28).
Fig. 5.
Sodium carbonate is more detrimental to flies
’
ability to escape
compared with other salts. For each concentration of each chemical tested
we briefly anesthetized 20
Drosophila virilis
with CO
2
and sprinkled them
onto a 400-mL jar (44-cm
2
surface area). Fifteen minutes later we scored each
fly for having escaped (0) or being trapped (1). Chemicals aside from MLW
were tested at identical molarities. Shading indicates bootstrapped 95%
confidence intervals.
13486
|
www.pnas.org/cgi/doi/10.1073/pnas.1714874114
van Breugel and Dickinson
that the three which had a deleterious effect all contained
dimethicone (Fig. 6
E
), which was absent from the three neutral
brands (
Fig. S3
C
). We repeated the dunk assay with pure water
after applying a surface film of 0, 2, or 10
μ
L of dimethicone
(viscosity 5 cSt; Sigma-Aldrich). As little as 45 nL of dimethicone
per cm
2
of water was enough to trap 50% of the flies (Fig. 6
F
).
Other artificial polymers such as trimethylsiloxysilicate and
vp/hexadecene copolymer are likely also problematic (
Fig. S3
C
).
Discussion
Through a series of experiments with solutions varying in salinity,
pH, and charge density, we showed that a high concentration of
Na
2
CO
3
makes it more difficult for alkali flies to escape from the
surface of water by facilitating the penetration of water into the
air pockets between individual hairs. The effect of Na
2
CO
3
is
surprising, considering that, like most salts, Na
2
CO
3
increases
the surface tension (
+
2% for 1 M solution) (20) as well as the
density of water. Both effects should theoretically increase the
recovered work by making large bubbles more stable and pro-
viding a larger buoyancy force. The fact that other salts, including
K
2
CO
3
, have a significantly smaller effect demonstrates that this
phenomenon involves interactions of specific ion pairs, and not with
Debye
–
Hückel or Derjaguin
–
Landau
–
Verwey
–
Overbeek models.
Instead, we offer a model based on the Hofmeister series to
explain this phenomenon, which we term
“
ion-facilitated wetting.
”
In 1888, Hofmeister (21) discovered that certain ions are more
likely to precipitate proteins out of egg white. The same order of
compounds
—
now known as the Hofmeister series
—
explains a
wide range of phenomenon including solubility and surface
tension (22). The order of anions is
ClO
-
4
<
I
-
<
Cl
-
<
F
-
<
OH
-
<
PO
3
-
4
<
SO
2
-
4
<
CO
2
-
3
.
Ions to the right of Cl
−
(kosmotropes) attract large hydration
shells that structure the surrounding water, thereby increasing
surface tension and driving ions away from the air
–
water inter-
face. Ions to the left of Cl
−
(chaotropes) have the opposite effect;
they tend to accumulate at the surface of an air
–
water interface
(23). Cations are arranged in following order:
NH
+
4
<
K
+
<
Na
+
<
Li
+
<
Mg
2
+
<
Ca
2
+
.
Again, ions on the right have larger hydration shells, although
anions generally have a more pronounced effect than cations.
The underlying principles that give rise to the Hofmeister series
are not well understood; however, the sequence is generally
correlated with the ratio of ionic charge to ionic radii (Fig. 7
A
)
(values from ref. 24).
Our result that Na
2
CO
3
,butnotK
2
CO
3
, has a strong effect on
wetting suggests that ion-facilitated wetting is dependent on the
precise combination of cations and anions. Typically, Hofmeister
effects of anions and cations are considered to be independent of
one another (25); however, some phenomena are known to de-
pend on ion specific pairs, such as the inhibition of bubble co-
alescence (26, 27). The physical basis of ion-facilitated wetting and
bubble coalescence are likely related, because both have macro-
scale consequences similar to those caused by surfactants (which
increase wetting and decrease bubble coalescence), yet they
operate through a completely different mechanism. However, our
results cannot be explained by the current models of bubble co-
alescence. We propose that the relative distance of the cations and
anions from the air
–
water interface plays a crucial role. Fig. 7
B
illustrates how, as a rough approximation, CO
3
2
−
is on average
situated 0.04 nm closer to the surface than Na
+
in Na
2
CO
3
so-
lution. This suggests that the carbonate ions will have a larger
influence on the surface characteristics in the presence of Na
+
,
giving the air
–
water interface a slight negative charge. Taking this
relative distance as well as the ion charge density into account
explains 77% of the variance in the correlation with the likelihood
of flies
’
becoming trapped at the surface (Fig. 7
C
). In contrast, in a
K
2
CO
3
solution the ions are nearly equidistant from the surface.
According to our theory, CaCO
3
,MgCO
3
, and Li
2
CO
3
would
have even stronger effects on wetting, as the hydration shells of
these cations are even larger. However, these salts are only soluble
Fig. 6.
Oils, notably dimethicone (a common ingredient in sunscreens and
cosmetics), annihilates alkali flies
’
superhydrophobic properties. (
A
)Force
traces of five flies dipped into MLW (blue), as in Fig. 2
A
, and MLW containing
dissolved Neutrogena Ultra Sheer SPF 50 sunscreen (brown). To prepare the
solution, 160 mg of sunscreen (
B
) was rubbed into both hands and allowed to
set for 15 min. We then poured 300 mL of MLW over one hand, and used the
run-off solution. (
B
) 160 mg sunscreen. (
C
) Work done on flies to escape MLW
or pure water, with or without Neutrogena Ultrasheer SPF 50 sunscreen (NS)
run-off. (
i
) Sunscreen set for 5 min before rinsing with MLW. (
ii
) Hands
thoroughly washed with soap and rinsed with warm water before rinsing.
(
iii
) Sunscreen set for 5 min before rinsing with pure water. (
iv
) Sunscreen set
for 15 min before rinsing with MLW. (
D
) Fraction of flies stuck to surface when
dunked into pure water with increasing concentrations of Neutrogena sun-
screen.
n
=
30 flies for each condition; mean and 95% confidence intervals are
shown. (
E
) Fraction of flies stuck when dunked into pure water with different
types of sunscreen (20 mg each). Same procedure as
D
.(
F
) To test whether
dimethicone is sufficient to trap flies on the water
’
s surface we applied 0, 2, or
10
μ
L to the surface and performed the same test described in
D
.
Fig. 7.
Ion-specific interactions including relative hydration shell size and
charge density help to explain ion-facilitated wetting. (
A
) Hofmeister series
of anions is correlated with the ratio of ion charge to radius. Black line shows
the regression (
P
=
0.004,
r
2
=
0.77). Diagrams depict ions (red) and their
hydration shells (blue), drawn to scale using the values reported in ref. 24.
(
B
) Comparison of Na
+
and K
+
ions to CO
3
2
−
.(
C
) Correlation of the fraction
of flies stuck in solutions from Fig. 5 (mean across concentrations) with the
product of the relative size of the cation and anion hydration shells and the
ratio of the charge (q) and radius (r) of the ion closest to the air
–
water in-
terface. Black line shows the regression (
P
=
0.024,
r
2
=
0.77).
van Breugel and Dickinson
PNAS
|
December 19, 2017
|
vol. 114
|
no. 51
|
13487
ECOLOGY
CHEMISTRY
at exceptionally low concentrations (0.1 mM to 0.17 M), not the
∼
0.5 M concentrations that are necessary, which makes Na
2
CO
3
the most potent compound for ion-facilitated wetting.
The physical mechanism by which the slight negative charge at
the air
–
water interface might increase the likelihood of wetting is
not immediately clear. One possible explanation involves elec-
trostatic attraction between the flies
’
surface and the negatively
charged fluid layer. Recent research has shown that Cassie
–
Baxter-to-Wenzel-state transitions are more likely to occur in the
presence of an applied voltage, which causes the formation of an
electric double layer (28
–
30). The electric double layer increases
the attraction between the interfacial water and individual
roughness elements on the surface, thereby pulling the solution
into the gaps and facilitating the transition to the wetted state. In
experiments in which the distance between roughness elements
was 4
μ
m (alkali flies
’
hairs are 3.2
μ
m apart), a voltage of 22 V
was required to cause wetting (30). To relate these experiments
with our results, we performed a rou
gh calculation to determine the
molar concentration of Na
2
CO
3
needed to generate a 22-V po-
tential between the water surface and the flies
’
cuticle (
SI Methods
).
Our model suggests that a molarity of
∼
0.15 M of Na
2
CO
3
is
necessary to induce wetting, which is within a factor of 4 of the
molarity we observed as necessary in our experiments, suggesting
that this is a plausible mechanism warranting further study. A
more thorough investigation would require detailed simulations
of molecular dynamics that are beyond the scope of this paper.
Our theory also explains the role of the cuticular hydrocarbons
in preventing wetting. The cuticle underneath the hydrocarbon
layer is largely composed of chitin, which is slightly polar [static
dielectric permittivity
=
15 (31)]. Thus, the nonpolar hydrocarbon
layer [static dielectric permittivity
=
∼
2 (32)] helps to insulate the
chitin surface from the electric double layer, reducing the likeli-
hood of wetting. This theory is consistent with our finding that
cuticular hydrocarbons do not influence wetting in pure water, as
there would be no electric double layer.
Compared with the six other species we investigated,
E. hians
were the only species that resisted wetting in the presence of
Na
2
CO
3
, an adaptation that allows them to occupy a rare but
ecologically important niche. Remarkably, the trait that allows
them to forage and lay eggs in such an extreme aquatic envi-
ronment arises from just a few minor changes in physical
and chemical properties. These adaptations likely evolved
over time in response to the slowly increasing concentration of
mineral salts (such as Na
2
CO
3
) in alkaline lakes across the
world. In recent times, the selective pressures on the alkali flies
at Mono Lake have become even stronger. Between 1941 and
1982 the concentration of mineral salts in the lake doubled as a
result of Los Angeles
’
policy of diverting water from the Eastern
Sierra. Our experiments, however, suggest that this increase in
ion concentration has had only a small influence on the flies
’
ability to dive and resurface in the lake. By comparison, the in-
creasing salinity has had a much larger detrimental effect on the
flies
’
larvae (33).
The most important adaptation that made the niche of un-
derwater feeding available to the alkali fly, however, was not a
physical or chemical one. Rather, it was the behavioral urge to
crawl under water and forage in the first place. We suspect that
their ancestors evolved this unusual behavior in lean times, when
surface food was a limiting resource but underwater algae were
abundant. In addition, selection against underwater foraging pre-
sumably decreased in alkaline lakes, because the caustic chemistry
makes them uninhabitable for fish.
ACKNOWLEDGMENTS.
We thank Jocelyn Millar, who generously performed
the GC-MS analysis in this paper and provided valuable feedback on
hydrocarbon chemistry; Rob Phillips, who also provided helpful comments
on the manuscript; Dave Marquart, who helped procure necessary permits;
Aisling Farrell, who helped collect
Helaeomyia petrolei
from the La Brea Tar
Pits; and Victoria Orphan and Sean Mullin, who helped prepare SEM speci-
mens. This work was supported by the National Geographic Society
’
s Com-
mittee for Research and Exploration, Grant 9645-15.
1. Twain M (1872)
Roughing It
(American Publishing Company, Hartford, CT).
2. Hart J (1996)
Storm over Mono: The Mono Lake Battle and the California Water
Future
(Univ of California Press, Berkeley, CA).
3. Tierney T (2000)
Geology of the Mono Basin
(Kutsavi Press/Mono Lake Committee,
Lee Vining, CA).
4. Vorster P (1985) A water balance forecast model for Mono Lake, California. PhD dis-
sertation (California State Univ, Hayward, CA). Available at www.monobasinresearch.
org/onlinereports/waterbalance.php. Accessed July 10, 2016.
5. Gao X, Jiang L (2004) Biophysics: Water-repellent legs of water striders.
Nature
432:36.
6. Flynn MR, Bush JWM (2008) Underwater breathing: The mechanics of plastron res-
piration.
J Fluid Mech
608:275
–
296.
7. Foley C, White B (1989) Occurence of Ephydra hians Say (Diptera: Ephydridae) in deep
water in Mono Lake, California.
Bull South Calif Acad Sci
88:4041.
8. Adam NK (1948) Principles of penetration of liquids into solids.
Discuss Faraday Soc
3:
5
–
11.
9. Holdgate MW (1955) The wetting of insect cuticles by water.
J Exp Biol
32:591
–
617.
10. Cassie ABD, Baxter S (1944) Wettability of porous surfaces.
Trans Faraday Soc
40:
546
–
551.
11. Barthlott W, Ehler N (1977) Raster-Elektronenmikroskopie der Epidermis-Oberflächen
von Spermatophyten (Scanning electron microscopy of epidermal surfaces of Sper-
matophyta).
Trop und Subtrop Pflanzenwelt
19:110.
12. Wenzel RN (1936) Resistance of solid surfaces to wetting by water.
J Ind Eng Chem
28:
988
–
994.
13. Feng BL, et al. (2002) Super-hydrophobic surfaces: From natural to artificial.
Adv
Mater
14:1857
–
1860.
14. Yuan Y, Lee TR (2013) Contact angle and wetting properties.
Surface Science
Techniques
, eds Bracco G, Holst B (Springer, Berlin), pp 3
–
34.
15. Jones & Stokes Associates Inc. (1993) Environmental impact report for the review of
Mono Basin water rights of the city of Los Angeles (Jones & Stokes Associates Inc.,
Sacramento, CA).
16. Denny MW (1993)
Air and Water: The Biology and Physics of Life
’
s Media
(Princeton
Univ, Princeton).
17. Morgan J, Townley S, Kemble G, Smith R (2002) Measurement of physical and me-
chanical properties of beeswax.
Mater Sci Technol
18:463
–
467.
18. Shi B, Zhao S, Jia L, Wang L (2007) Surface characterization of chitin by inverse gas
chromatography.
Carbohydr Polym
67:398
–
402.
19. Young T (1805) An essay on the cohesion of fluids.
Philos Trans R Soc Lond
95:65
–
87.
20. Ozdemir O, Karakashev SI, Nguyen AV, Miller JD (2006) Adsorption of carbonate and
bicarbonate salts at the air
–
brine interface.
Int J Miner Process
81:149
–
158.
21. Hofmeister F (1888) Zur lehre von der wirkung der salze.
Arch für Exp Pathol und
Pharmakologie
24:247
–
260.
22. Zhang Y, Cremer PS (2006) Interactions between macromolecules and ions: The
Hofmeister series.
Curr Opin Chem Biol
10:658
–
663.
23. dos Santos AP, Diehl A, Levin Y (2010) Surface tensions, surface potentials, and the
Hofmeister series of electrolyte solutions.
Langmuir
26:10778
–
10783.
24. Marcus Y (1991) Thermodynamics of solvation of ions. Part 5. Gibbs free energy of
hydration at 298.15 K.
J Chem Soc Faraday Trans
87:2995
–
2999.
25. Lo Nostro P, Ninham BW (2012) Hofmeister phenomena: An update on ion specificity
in biology.
Chem Rev
112:2286
–
2322.
26. Craig VSJ, Ninham BW, Pashley RM (1993) Effect of electrolytes on bubble co-
alescence.
Nature
364:317
–
318.
27. Henry CL, Craig VSJ (2010) The link between ion specific bubble coalescence and
Hofmeister effects is the partitioning of ions within the interface.
Langmuir
26:
6478
–
6483.
28. Das S, Mitra SK, Chakraborty S (2012) Wenzel and Cassie-Baxter states of an elec-
trolytic drop on charged surfaces.
Phys Rev E Stat Nonlin Soft Matter Phys
86:011603.
29. Manukyan G, Oh JM, van den Ende D, Lammertink RGH, Mugele F (2011) Electrical
switching of wetting states on superhydrophobic surfaces: A route towards reversible
Cassie-to-Wenzel transitions.
Phys Rev Lett
106:014501.
30. Krupenkin TN, Taylor JA, Schneider TM, Yang S (2004) From rolling ball to complete
wetting: The dynamic tuning of liquids on nanostructured surfaces.
Langmuir
20:
3824
–
3827.
31. Seoudi R, Nada AMA (2007) Molecular structure and dielectric properties studies of
chitin and its treated by acid, base and hypochlorite.
Carbohydr Polym
68:728
–
733.
32. Israelachvili JN (2011)
Intermolecular and Surface Forces
(Academic, Waltham, MA),
3rd Ed.
33. Herbst DB, Conte FP, Brookes VJ (1988) Osmoregulation in an alkaline salt lake insect,
Ephydra (hydropyrus) hians Say (Diptera: Ephydridae) in relation to water chemistry.
J Insect Physiol
34:903
–
909.
13488
|
www.pnas.org/cgi/doi/10.1073/pnas.1714874114
van Breugel and Dickinson